J. A@.
la32
Food chem. 1993, 41, 1332-1336
Detection of Dimetridazole and Other Nitroimidazole Residues in Turkey Using an Immunoassay Larry H. Stanker;J Cathy McKeown,? Bruce E. Watkins? Martin Vanderlaan,+Richard Ellis,l and Jess Rajani Biomedical Sciences Division and Chemical Sciences Division, Lawrence Livermore National Laboratory, University of California, P.O.Box 5507 L-452,Livermore, California 94550,and Food Safety Inspection Service, U.S.Department of Agriculture, Washington, D.C. 20250 A series of monoclonal antibodies were generated that can bind dimetridazole, a nitroimidazole drug used in veterinary medicine. A competition enzyme-linked immunosorption assay (cELISA) is described and is used to characterize the binding of these antibodies to a number of nitroimidazole drugs. The 50% inhibition of control occurred for the most sensitive of these antibodies a t 12ng/mL for dimetridazole, 75 ng/mL for hydroxydimetridazole, 25 ng/mL for ipronidazole, 2000 ng/mL for hydroxyipronidazole, and 20 000 ng/mL for metronidazole. An extraction method for these nitroimidazoles that is compatible with the immunoassay is described. Using this method, as little as 1 ng of dimetridazole could be detected in turkey muscle.
INTRODUCTION Dimetridazole (DMZ) is a nitroimidazole compound used in veterinary medicine to treat poultry for coccidiosis and histomoniasis ("black head"). The major metabolite of DMZ is 2-(hydroxymethyl)-l-methyl-5-nitroimidazole (DMZOH),formed by hydroxylation of the 2-methyl group. Further oxidation to the 2-carboxylicacid analog has been reported (Law et al., 1963). Dimetridazole and ipronidazole (IPR) were approved by the U.S. Food and Drug Administration (FDA) in 1971 and 1965,respectively, for use in turkeys. Both drugs have how been withdrawn from the market, DMZ in 1987 and IPR in 1989. Many countries, however, including Canada, Australia, and Denmark permit use of these drugs in food animals. Thus, a sensitive rapid screening method for detection of these compounds in food products is desirable. In this paper we describe the development of a monoclonal antibody to DMZ and ita application to detection of nitroimidazoles in turkey. A number of analytical methods have been described for determination of nitroimidazoles. A tandem mass spectrometry method to detect DMZ, IPR, and their alcohol metabolites in the low parta per billion (ppb) range has recently been described by Matusik et al. (1992). Mallison and Henry (1992) have described a liquid chromatographic method that can measure DMZOH with a 2 ppb detection limit. Their method also can detect the hydroxyl metabolite of ipronidazole (IPROH) as well as the parent compounds in swine and turkey muscle tissue. Carignon et al. (1988)have reported a high-pressure liquid chromatographic (HPLC) method to measure DMZ in pork tissue. Gas chromatographic methods for detection of DMZOH in swine muscle have been reported by Newkirk et al. (1990)and by Stone et al. (1978). A polarographic method with a 2 ppb limit of sensitivity for DMZOH in swine muscle tissue has been described by Craine et al.
* Address correspondence to this author at Food Animal Protection Research Laboratory, USDA-ARS, Route 5, Box 810,College Station, TX 77845. t Biomedical Sciences Division. t Chemical Sciences Division. 8 Food Safety Inspection Service. 002 1-856 lI93l l~l-l332$O4.OO/O
(1974). Garland et al. (1980)described a negative ion chemical ionization method for IPR and IPROH. Both IPR and DMZ were recovered from swine feed and analyzed by Roybal et al. (1987)using an HPLC method, and MacDonald et al. (1971)developed a method to detect IPR in turkey tissue at the 2 ppb level. A competition enzyme-linked immunosorbent assay (cELISA) is described in this paper. The cELISA is based on a monoclonal antibody and detects DZM, DMZOH, IPR, and IPROH. Turkey muscle tissues fortified with DMZ, DMZOH, and IPR a t levels ranging from 0.3 to 50 ppb were evaluated with the cELISA. MATERIALS AND METHODS Reagents. Dimetridazole (DMZ) (1,2-dimethyl-&nitroimidazole), hydroxydimetridazole (DMZOH) [l-methyl-a-(hydroxymethyl)-5-nitroimidazolel,ipronidazole (IPR) (l-methyl2-isopropyl-5-nitroimidazole),and metronidazole (METRON) [1-(hydroxyethyl)-2-methyl-5-nitroimidazole] were all provided by the U.S. Department of Agriculture, Food Safety Inspection Services. The above chemicals were greater than 95% pure for use as analytical standards. Hapten Preparation. The immunogen was prepared as follows. Hydroxydimetridazole was reacted with succinic anhydride to form the hemisuccinate and then conjugated to keyhole limpet hemocyanin (KLH) (DMZ-KLH) and to bovine serum albumin (BSA)(DMZ-BSA)using the mixed anhydridemethod (Erlanger et al., 1969). Monoclonal Antibody Production. Six-month-old BALB/ cBkl mice (Bantin and Kmgman Laboratories, Fremont, CA) were injected intraperitoneally (ip) with 100 pg of the DMZKLH conjugatemixed 1:lwith completeFreund's adjuvant.Mice received a single ip injection every other week for a totalof three injections. Four days prior to fusion, each m o w was given an intrasplenic injection of 100pg of DMZ-BSA conjugate in sterile saline. The spleen was removed, and the splenocytm were fused with SP2/0 myeloma celle and grown under conditions described by Stanker et al. (1986). A direct-binding ELISA, described by Stanker et al. (1986) and modified as described below, was used to screen culture the fluids from the growing hybridomas for antibodiea to DMZ. DMZOH linked to BSA (DMZrBSA) served as antigen in both the direct-bindingELISA and the competition ELISA (cELISA)
(describedbelow). The coating antigen WBB prepared as follows. Fifty milligrams of the BSA dissolved in 5 mL of distilled water was added to 60 mg of the hemieuccinate. The pH was adjusted to 7.0 by addition of 1 N sodium hydroxide. Then, 60 mg of
Q 1909 American Chemical Soci8W
J. A@.
Immunoassay Detection of Dimetrldazole in Turkey
l-ethyl-3-[3-(dimethylamino)propyl] carbodiimide (EDC) (Pierce Chemical Co., Rockford, IL) was added, and the mixture was stirred overnight at ambient temperature and dialyzed for 48 h against four changes of PBS (0.01 M sodium phosphate, 0.15 M NaC1, pH 7.2). Microtiter plates (Nunc Maxisorb, Denmark) were coated with DMZ-BSA by addition of 100 pL/well of a 200 ng/mL solution of DMZ-BSA in carbonate buffer. The DMZBSA was incubated in uncovered plates at 37 O C for 18 h to evaporate the liquid and allow the DMZ-BSA to coat the bottom of the microtiter wells. The "coated" plates were then stored in sealed plastic bags at 4 O C and used within 2 weeks. Nonreacted sites on the plastic microtiter plates were blocked by adding 400 pL of assay buffer (AB) [0.1M Tris, 0.15 M NaC1,0.001% Tween 20,0.005% Thimerosal, and calf serum (1mL/lOO mL of AB), pH 7.21 and then incubating the plates at room temperature for 1h. The blocking solution was removed, 50 pL of the hybridoma supernatant(8) or the anti-dimetridazole Mab was added, and the plates were incubated for 1 h at 37 OC. The plates were carefully washed with a solution of 0.05% Tween 20 in water, and peroxidase conjugated goat anti-mouse antiserum (Sigma Chemical Co., St. Louis, MO) diluted 1:500 in AB was added to each well. Following a second l-h incubation at 37 "C, the plates were washed again and the substrate 2,2'-azinobis(3-ethylbenzthiazolinesulfonic acid) (ABTS) was added. Absorbance measurements at 405 nm were taken and the resulting data transferred to a Macintosh computer and analyzed with the Cyberdoma ELISA software described by Slezak et al. (1983). Hybridoma cells from wells showinga positive response in the ELISA screen were expanded and subcloned twice by limiting dilution to ensure their monoclonal origin. Ascites fluid was prepared in irradiated mice (Stanker et al., 1986), and the monoclonal antibodies were purified from the ascites by hydroxyapatite chromatography (Stanker et al., 1985). Isotype determination was done by ELISA using mouse heavy- and lightchain-specific antisera (Southern Biotechnology Associates, Birmingham, AL). Competition ELISA (cELISA)for Dimetridazole. A cELISA for dimetridazole and other related nitroimidazoleswas as follows. To each well of an antigen-coated, assay-bufferblocked, microtiter plate was added 100 pL of sample buffer (SB) [0.5 M Tris, 0.15 M NaCl, 0.001% Tween 20, 0.005% thimerosal, and calf serum (1 mL/100 mL AB), pH 7.21. Competitors dissolved in 100 pL of SB were added to the microassay plate and serially diluted (a 2-fold series). Next, 100 pL of SB containing a predetermined amount of anti-dimetridazole monoclonal antibody (for Mab Di-7,100 A of a 300 ng/mL solution)was added to each well. Thus, each well contained 200 pL of sample buffer containing antibody and competitor. The amount of anti-dimetridazole antibody used is approximately that concentration of antibody that results in 50% of maximum activityin a direct bindingELISA where no computerwas present. The plates were sealed with plastic wrap, incubated for 1 h at 37 "C, and then processed withperoxidase-conjugatedanti-mouse antibody (diluted 1/50) and substrate as described above. In eachexperiment,microtiter wells containingall components except competitor were prepared and the activity in theee wells was taken to represent 100% activity (control web). The test wells, each containing different amounts of competitor, were normalized to the 100% activity wells, and percent inhibition was calculated as follows. % inhibition of control = 11- (OD, of test)/(ODm,
of no competitor wells)] x 100
Dimetridazole Extraction. Dimetridazole and other nitroimidazoles were extracted from turkey as follows: A 50-g sample of turkey breast (purchased from local supermarkets) was placed in a 2 " L conical polypropylene centrifuge tube; were added, followed by 90mL 20 g of NaCl and 20 g of KzHPO~ of ethyl acetate. The sample was then homogenized in a polytron (Brinkman Instruments, Westbury, NY)at medium-high speed (setting 6) for 1.5 min. The resulting slurry was centrifuged at lOOOg for 2 min in a countertop centrifuge to sediment the particulate. The particulate was subjected to two additional extractions (90mL of ethyl acetate was used for each extraction). The ethyl acetate fractions from each extraction were pooled
~ ~ ~ d ~ h vol. e m41, . , NO. S , I Q Q ~
iaaa
(approximately270mL) into a separatoryfunnel. Distilled water (5 mL) was then added and the sample shaken by hand for 30 8. Next, 5 mL of 1 N HC1 was added and the sample shaken for an additional 30 8. The phases were allowed to separate, and the aqueous fraction (bottom) was collected into a 50-mL centrifuge tube. Three additional extractions each using 6 mL of 1N HCl were performed and the aqueous fractions collected and pooled. The pooled aqueous fraction was cooled on ice, 10 g of KzHPO, added, and the tube again allowed to cool on ice. An additional 10 g of K*HP04was then added and the sample cooled (not all of the salt is dissolved at thistime). Methylene chloride (10 mL) was then added and the tube capped, shaken, and centrifuged for 1 min at approximately 1OOOg in a tabletop centrifuge to separate the phases. The methylene chloride fraction (top) was recovered to a separatory funnel. Three additional methylene chloride extractions were performed. Next, 10 mL of alkaline wash buffer (0.1M sodium phosphate, 0.15 M NaC1, pH 8.0) was added to the pooled methylene chloride fraction and shaken for 10 8. The phases were allowed to separate, and the aqueous phase was discarded. The methylene chloride fraction was partitioned a second time against alkaline wash buffer and evaporated to dryness at room temperature under a stream of nitrogen. Prolonged drying was avoided since these nitroimidazoles can be volatilized. Finally, the residue was dissolved in 1 mL of sample buffer and used in the cELISA. Turkey breast samples were spiked with nitroimidazole standards, in assay buffer, by delivering between 100 and 500 pL of standard to a coarsely ground meat sample at three or four locations. Assay buffer alone was added to the unspiked samples. RESULTS
Hybridoma Production. Spleen cells from a BALB/c mouse immunized with DMZ-KLH were fused with SP 210 myeloma cells, and the resulting hybridomas were cultured in 30 96-well microculture dishes. Greater than 90% of the wells had growing hybridomas at 10 days after the fusion. Those wells containing anti-dimetridazole antibodies were identified by evaluating the supernatant from each well for antibodies reactive to DMZ-BSA in an ELISA. (BSA represents an extraneous protein because the immunogen was DMZ-KLH; therefore, anti-KLH antibodies were not detected.) Approximately 115 wells were observed to be secreting antibody that recognized the DMZ-BSA conjugate but not BSA itself. Cells from these wells were expanded and tested against DMZ-BSA, DMZ-KLH, BSA, and KLH. Forty-eight wells were observed to contain cells that secreted antibody that recognized both hapten conjugatesbut did not bind either of the unconjugated carrier proteins. These were next evaluated for their abilityto produce antibodiesthat bound dimetridazole in the cELISA. Only 11wells were observed to produce antibodies that bound dimetridazole in the cELISA. These 11were subcloned twice to ensure their monoclonal origin, and their isotypes were determined. The antibodies, referred to as Di-1-Di-11, were found to have either IgM, IgG2a, or IgG2b heavy chains, and all had kappa light chains (Table I). Antibody Characterization. The amount of dimetridazole necessary to cause 50% inhibition of antibody binding (the ICw) for each of the 11Mabs isolated is shown in Table I. The ICw ranged from 2 to 260 ng, indicating the range in relative affinity that these antibodies have for dimetridazole. Typical cELISA data obtained with Mab Di-7 and DMZ, DMZOH, IPR,IPROH, and metronidazole are shown in Figure 1. Di-7 differentially binds these compounds and serves as an example for the other Mabs. Di-7 had the greatest sensitivity for DMZ (ICs0 of 1.0 f 0.6 ng) followed by IPR (ICm of 2.5 f 1.0 ng). There was approximately a 6-fold reduction in sensitivity to DMZOH. A 166-fold reduction in binding to the hydroxyl metabolite of IPR was observed. Met-
Stanker et al.
1934 J. A M . Food W m . ,Vol. 41, No. 8, 1993
' 7
Table I. IC~SValuer (NanoBrunr Der Well)* antibody i e o t y p e b DMZ DMZOH IPR IRPOH METRON 11 11 390 >1m 3 IgM Di-1 160 50 5ooo >loo00 18 IgM Di-2 Di-3 IgM 18 160 150 >loo00 >loo00 190 60 6o00 >loo00 IgM 120 Di-4 Di-6 IgM 200 1000 350 >SO00 >loo00 8Ooo >loo00 Di-6 I&zb 200 >loo0 >I000 IgGh 1.2 7.5 2.5 200 2000 Di-7 6OOO loo00 Di-8 IgGM 20 100 18 4600 loo00 Di-9 260 >I000 6OOO 250 80 loo00 Di-10 IgGa 100 35 >loo00 loo00 90 30 Di-11 IgM a ng/well is converted to ppb simply by multiplying by 10. All utilize kappa light chaina. > indicates that no inhibition waa detected at the amount indicated. ~
~~
6-
5h
P
Y
S
p Z
Y
4-
' 3-
' 21-
*
100
0
200
300
Antibody Concentration (ng/wrll)
Figure 2. Effects of antigen concentration and antibody concentration on the sensitivity of the cELISA for dimetridazole. Squares represent coating antigen at 100 nglwell, triangles 25 ng/well, and circles 10 ng well. 100 5 80 h
c
f 60 0
c 0
E 40
5
P c c
Competitor (ng/well)
Figure 1. Typical cELISA showing the ability of Mab Di-7 to bind DMZ (open circles), IPRON (solid circles), DMZOH (open squares), IPRONOH (solid squares), and METRON (open triangles). These reactions were in 100-pL volumes.
ronidazole was the most poorly bound nitroimidazole teated, resulting in an ICs0 at least 1666-foldless sensitive than DMZ. The other 10 Mabs had relative affinities similar to that of Di-7 but with lower absolute binding. These data (Table I) suggest that the relative rankings of the compounds by these antibodies are similar but that the relative affinities of the antibodies vary. Monoclonal antibody Di-7 was chosen for further study since it is an IgG immunoglobin and has the greatest relative sensitivity for dimetridazole. Di-7 was produced as an ascites fluid, purified as described, and stored at -20 OC until used. Competition ELISA Characterization. The sensitivity of a cELISA is influenced by the amounts of both the primary antibody and the coating antigen (Pesce et al., 1981). Therefore, the amounts of both the antidimetridazoleantibody (Di-7)and the coating antigen were varied in an effort to determine the optimum conditions for the assay. In these experiments the dimetridazole coating antigen, DMZ-BSA, was used at three levels: 100, 25, and 10 ng/well. At each coating antigen level cELISAs were performed using the Di-7 antibody at concentrations of 240,120,60, and 30 ng/well. Coating antigen levels of 6 and 3 ng/well also were evaluated but gave minimal signals that were difficult to detect over background. Antibody concentrations below 30 ng/well were not evaluated. In these experiments, the peroxidase-conjugated antibody was used at a 1/100 dilution of the stock material as supplied by the manufacturer. As expected, sensitivity was increased by reducing the level of antigen used to coat the microtiter wells and by reducing the Di-7 antibody concentration. The resulta from these experiments are summarized in Figure 2. To obtain the
ae 20 0
+.
10 100 Dimetrldazo le Competitor (nglwell) 1
Figure 3. Competition curve obtained with dimetridazole standard and monoclonal antibody Di-7. Bars = *standard deviation (N= 9).
sensitivity necessary to measure dimetridazole near the tolerance level (2 ppb) in all subsequent cELISA experiments, the Di-7 antibody was used at 30 ng/well with a coating antigen concentration of 20 ng/well. Figure 3 shows typical cELISA curves obtained with the DMZ standard under these conditions (bars = standard deviation). The relatively large standard deviation observed in the experiments shown in Figure 3 most likely reflects the need to run the assay under conditions that result in maximum sensitivity (i.e., detection in the nanogram range). Detection in Turkey Muscle. To detect DMZ and other nitroimidazole compounds in turkey, the analyte must be extracted and presented to the antibody in a medium compatible with antibody function. Briefly, 50 g of ground turkey breast muscle was fortified with various levels of DMZ, DMZOH, and IPR as described above. The sample was then extracted with ethyl acetate. Analysis of the extracts, following dry-down and resuspension into buffer, by cELISA resulted in unacceptably high background levels in both control and spiked samples. It was necessary to further purify the samples by extraction of the ethyl acetate fraction against 1M aqueoushydrochloric acid, neutralize the acid, and then extract with methylene chloride. Samples from the methylene chloride fraction were dried and resuspended in methanol for analysis in the cELISA. Again, control turkey samples gave unacceptable levels of inhibition in the cELISA. Inhibitions as high as 40% of control were often observed (data not
I"wx#asay
Detectton of Mmetrklazok In Turkey
120
J. A@.
Food-.,
Vol. 41, No. 8, 1993 lW5
I
DYZOH
DYZ
CH
I 3
-1
1
10
IPR
CHCHOH
CH3
I 2
2
100 I PROH
Spi Ice (ppb)
Figure 4. Inhibition curves obtained in fortified turkey breast muscle. Sampleswere fortified as indicated and then extracted. (Triangles) DMZ (open circles) IPR (solid circles) DMZOH. Bare = &standarddeviation (N1 3).
shown). The methylene chloride fraction was next washed by partitioning against alkaline buffer. Finally, the alkaline-washedmethylene chloridefraction was dried and the residue resuspended in methanol and analyzed with the cELISA. Nonspiked control turkey samples gave background levels of inhibition between 10% and 20% of control. These levels were sufficiently low to allow detection of dimetridazole samples spiked at parta per billion levels. Figure 4 summarizesthe data obtained when ground turkey muscle samples were fortified before extraction with DMZ, DMZOH, and IPR. In each case the bars represent the standard deviation. DISCUSSION Dimetridazole is a small nitroimidazole molecule that must first be conjugated to a carrier molecule to render the dimetridazole immunogenic. The linkage strategy employed clearly resulted in a potent immunogen. The screening system used allowed quick elimination of the majority of the hybridomas obtained following a cell fusion from those that recognized dimetridazole but only if conjugated to a protein. The 115 hybridomas detected in the initial screen represented approximately 1-4% of the total number of hybridomas observed. This value is consistent with previous studies (Stanker et al., 1987,1989). However, only 11hybridomas (approximately0.2%) were observed to bind free dimetridazole. Analysis of the 11 monoclonals indicated that 7 were IgM antibodies. The ICM for DMZ ranged from 1.2 to 260 ng for these antibodies, implying a large range in relative affinity and suggesting that they represent multiple clones. Only 1 of the 11 antibodies, Di-7, had a relative affmity great enough to be useful for analyte detection in the parta per billion range. These data point out the necessity to screen large numbers of hybridomas to quickly detect the few monoclonals that bind the "free" hapten with the affmity desired. The high number of IgM antibodies and the low relative affinity of many of the IgG antibodies may reflect the limitations inherent in the rapid immunization schedule used. Analysis of the binding properties of the 11hybridomas to related nitroimidazole compounds suggests that the antibodies can be divided into at least two epitope groups. The first group is composed of Di-1-Di-4, Di-7, Di-8, Di10, and Di-ll. These antibodies generally Bind DMZ with the greatest relative affinity followed by DMZOH, IPR, and IPROH and with little or no binding to METRON. Within this group, however, there may be subgroupsbased on the ability of the antibodies to differentially bind
METRON
Figure 5. Chemicalstructures for DMZ,DMZOH,IPR, IPROH, and METRON. IPROH and METRON. The second group is composed of Di-5, Di-6, and Di-9. These antibodies have a lower relative affinity than those in the first group on the basis of their ICs0 values and are most specific for DMZ with little binding to the other compounds. The antibodies in thie latter group may simply represent antibodies derived from cells that were in different compartments of the affmity maturation process that occurs during an immune response. Nevertheless, the presence of different binding properties for these antibodies points out that even though these nitroimidazole compounds are small molecules (e.g., 141.13DaforDMZ),numerousantibcdiescanbegenerated in a single immune response, and that binding for each of these antibodies is influenced by different parameters of the molecule. Differential binding of monoclonal antibodies to small molecules also has been observed by others (Hill et al., 1991; Stanker et al., 1991; Newsome et al., 1990),and the results presented here point out the power of selecting a specific monoclonal antibody. The specificity of binding for the antibodies described here is difficult to elucidate, in part because of the limited number of related analytes that were available for study (Figure 5). Clearly, substitution of an ethanol group on the no. 1nitrogen as seen in METRON inhibits binding and suggests that thie position is important for antibody binding. However, it is not known if these antibodies interact directly with the no. 1nitrogen or if it is simply that substitution of an ethanol group on the no. 1nitrogen vs a methyl group alters the molecule in such a way that antibody binding is inhibited. Molecular modeling of DMZ, DMZOH, IPR, IPROH, and METRON was performed using the modeling programs interfaced by the CAche Molecular Modeling System (CAche-Tectronics, Beaverton, OR). These calculations suggest that the substitutions in IPROH and METRON facilitate rotation of the molecule around the no. 2 carbon, resulting in a significantlyaltered electron density. This might explain the differential antibody binding resulta. Regardless of the exact nature of the binding, the antibodies described here are able to bind, rather effectively, a related group of nitroimidazole compounds. The data shown here using fortified turkey muscle samples demonstrate that Mab Di-7 can be configured into an immunoassay capable of detecting these compounds. A clear dose-response curve was obtained in turkey muscle with the immunoassay,after extraction and cleanup of the material. Simple extractions in aqueous buffers or in ethyl acetate were unsuccessful;in both cases the extracts appeared to contain interfering substances that resulted in apparent competition (approaching40%) in control tissues. Further cleanup by extracting the sample against alkaline buffer was required to remove most of the interfering substances. The data clearly
1886 J. A*.
FoodChem., Vol. 41, No. 8, 1993
demonstrate that the cleanup methods coupled with the immunoassay described here is capable of detecting DMZ, DMZOH, and IPR in extracts of turkey muscle at low parts per billion sensitivity. The relative sensitivity of the immunoassay using fortified turkey breast samples parallels that seen with standards. T h e assay is most sensitive for DMZ and IPR and least sensitive for their alcohol metabolites. It is equally clear that the length of the cleanup method detracts from the usefulness of the immunoassay. Further studies are needed to streamline the cleanup method and reduce the use of organic solvents. ACKNOWLEDGMENT Funding for this work was provided by the U.S. Department of Agriculture, Food Safety Inspection Services, under Interagency Agreement 13-37-7-046,and the work was performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under Contract W-7405-ENG-48. LITERATURE CITED Carignan, G.; Macintosh, A. I.; Skakum, W.; Sved, S. Dimetridazole Residues in Pork Tissue. 11. Application of Liquid Chromatographic Method to Monitor Elimination of Drug and ita Major Metabolite. J. Assoc. Off.Anal. Chem. 1988,71, 1146-1149. Craine, E.M.; Parnell, M. J.; Stone, L. R. A Method for Analysis of Swine Tissue for the Primary Metabolite of Dimetridazole at the 2 ppb Level. J.Agric. Food Chem. 1974,22,877-881. Erlanger, B. F.; Borek, 0.F.; Beiser,S. M.; Leiberman, S. Steroidprotein Conjugates I1 Preparation and Characterization of Bovine Serum Albumin with Progesterone, Deoxycorticosterone, and Estrone. J. Biol. Chem. 1959,234,1090-1094. Garland, W.A.; Hodshon, B. J.; Chen, G.; W e h , G.; Felicitio, N. R.; MacDonald, A. Determination of Ipronidazole and its Principal Metabolite in Turkey Skin and Muscle by Combined Gas Chromatography-Negative Chemical Ionization Mass Spectrometry-Stable Isotope Dilution. J. Agric. Food Chem. 1980,28,273-277. Hill, A. S.; Mei, J. V.; Yin, C.; Ferguson, B. S.; Skerritt, J. H. Determination of the Ineect Growth Regulator Methoprene in Wheat Grain and Milliig Fractions Using an Enzyme Immunoassay. J. Agric. Food Chem. 1991,39,1882-1886. Law, G. L.; Manefield, G. P.; Muggleton, D. F.; Parnell, E. W. Dimetridazole: Absorption, Excretion and Metabolish in Turkeys. Nature 1963,197,1024. MacDonald, A.; Chen, G.; Kaykaty, M.; Fellig, J. Residue Analysis of Ipronidazole and ita Metabolite at the 2ppb Level in Turkey Tissue. J. Agric. Food Chem. 1971,19,1222-1227.
Stanker et ai.
Mallinson, E. T.; Henry, C.; Rowe, L. D. Determination of Nitroimidazole Metabolites in Swine and Turkey Muscle by Liquid Chromatography. J. AOAC Znt. 75,790-796. Matusik, J. E.; Leadbetter, M. G.; Barnes, C. J.; Sphon, A. Identificationof Dimetridazole, Ipronidazole,and their Alcohol Metabolites in Turkey Tissues by Thermospray Tandem M a Spectrometry. J. Agric. Food Chem. 1992,40,439-444. Newkirk, D. R.; Righter, H. F.; Schenck, F. J. Gas Chromatographic Determination of Incurred Dimetridazole Rseidues in Swine Tissues. J. Assoc. Off.Anal. Chem. 1990,73,702704. Newsome, W. H.; Colliis, P. G. Development of an ELISA for Urea Herbicides in Foods. Food Agric. Zmmunol. 1990,2,786 884. Pesce, A. J.; Ford, D.J.; Makler, M. T. Properties of Enzyme Immunoassays. In Enzyme Immunoassay; Iskikawa, E., Igaku S h o h Tokyo, 1981;pp 27Kawai, T., Miyai, K., Us.; 39. Roybal, J. E.; Munns, R. K.; Hurlbujt, J. A.; Shimoda, W.; Morison, T. R.; Vieira, C. L. Rapid Liquid Chromatographic Determination of Dimetridazole and Ipronidazole in Swine Feed. J. Assoc. Off.Anal. Chem. 1987, 70,626-630. S l e d , T.; Vanderlaan, M.; Jenaen, R. H. A Computer-based Data Analysis System for Enzyme-linked Immunoeorbent Assays. J. Zmmunol. Methods 1983,65,83-95. Stanker, L. H.; Vanderlaan, M.; Juarez-Salinas, H. One-step Purification of Mouse Monoclonal Antibodies from Ascites Fluid by Hydroxyapatite Chromatography. J. Immunol. Methods 1985, 76,157-169. Stanker, L. H.; Branscomb, E.; Vanderlaan, M.; Jensen, R. H. Monoclonal Antibodies Recognizing Single Amino Acid Substitution in Hemoglobin. J. Zmmunol. 1986,136, 6 1 6 622. Stanker, L. H.; Watkins, B.; Rogers, N.; Vanderlaan, M. Monoclonal Antibodies for Dioxin: Antibody Characterization and b y Development. Toxicology 1987,45,229-243. Stanker, L. H.; Van Emon, J.; Bigbee, C.; Watkins, B.; Morris, C.; Vanderlaan, M. An Immunoassay for Pyrethroids: Detaction of Permethrin in Meat. J. Agric. Food Chem. 1989, 37,834-839. Stanker, L. H.; Watkins, B.; Vanderlaan, M.; Ellis, R.; Rajan, J. Analysis of Heptachlor and Related Cyclodiene Insecticides in Food Products. In Zmmunoassays for Monitoring Human Exposure to Toxic Chemicals in Foods and the Environment; Vanderlaan, M., Stanker, L. H., Watkins, B., Roberta,D. W., Eds.;ACS Symposium Series 451;American Chemical Society: Washington, DC, 1991; pp 108-123. Stone, L. R.;Lawrence, J. W.; Hobson, D. L. GLC Determination in Swine Tisof l-methyl-2-hydroxymethyl-5-nitroimidazole sue. Res. Rep. LRS 1978,75 (Oct 2), 66. Received for review September 28, 1992. Revised manuscript received April 2, 1993. Accepted April 29, 1993.